How Tp Grow A Marketing Database

While the term database marketing has been around for at least a decade, its implementation has been less than successful in our industry. Before the Internet boom and increasingly sophisticated software that allows unlimited cross-referencing and on-the-fly data analysis, a database basically meant a relatively simple list of names, contact information, and buying history with the company that can be formatted for mailing purposes.

Over the past few years, however, databases have become extraordinarily robust marketing tools, providing companies with in-depth, up-to-date data that can be sliced and diced any which way for evaluation. While we may not need the level of complexity a Charles Schwab or a WalMart database requires, we do need one that meshes with our business and sales objectives. If we don't have up-to-the-minute information about our customers, we¼re not going to be able to sell to them effectively.

In the past, I've written of the importance of being able to integrate all disparate company databases - sales, production, accounting, etc. - into a single database. There are many reasons, but the most obvious is the elimination of duplicate, sometimes triplicate data entry (and greater likelihood of errors). A single database also ensures that collected data can be used throughout the organization by all departments. As I mentioned in last month's column, FileMaker Pro is proving its mettle as a foundation database since it accepts importation of a wide range of other database applications.


A viable marketing database, like a viable business, is continually evolving. Not to be viewed as a static object, a marketing database is an ongoing process of discovering information about customers and prospects, and utilizing that information to craft targeted, personalized messages. Hopefully, you will have already done a lot of this work, since it's not only critical to marketing efforts, but also central to the ongoing health of your organization.

Review. While this may seem unnecessary, it's a critical first step. Very much like a SWOT analysis, you should look at your company's existing business and marketing plans and objectives. Review current data structures and other information repositories, such as electronic job tickets. Evaluate how revenue is reported, how costs are applied and what reporting periods are useful. Your goal in this first phase is to gain a thorough understanding of your existing business model, what information you currently have that's useful and what augmentation you'll need to meet your objectives. You should also try to predict the roadblocks you'll likely encounter in developing your database and realizing your marketing objectives; e.g., tight economy, increased competition, inadequate staffing, etc. Opportunities should also be identified.

Determine data needs and who's to provide. While you may not be able to anticipate every piece of information you'd like to glean from your database, the more up-front preparation, the more time and money you'll save in the long run. Identify which files, fields, records, and backup documentation you want. While most databases are scalable, allowing you to add fields along the way, designing a structure at the outset that contains all key data fields gives all who will be providing data a roadmap to follow. This is also the time to create ID fields for segmentation; e.g., by SIC code, list type, market segment, etc. Think ahead. If you're likely to be purchasing lists from outside sources, SIC categorization will be important.

Load and test data. This stage ensures your data is accurate, complete, and has been imported into the database properly. Run tests, analyze, and generate reports to make sure the database is being built to meet your and other departmental needs. You want to catch mistakes at this point before the database has been completely built. Look for errors like misplaced decimal points, missing fields, inaccurate tables, etc.

Clean and clean again. While data hygiene is an ongoing maintenance essential, it's wise to start with a clean slate. Standardize your input format; e.g., upper/lower case for names, all caps for company name, etc. Conduct a thorough review of each record and correct misspellings, address omissions, and double-checking parsing of names and fields; e.g., first name and last name are in separate fields. If you will be overlaying data from outside sources, make sure your internal data is clean before doing a merge/purge. This will eliminate the cost of matching duplicate records.

Review and revise. Now that you've run some reports and analyzed records for accuracy, it's time to step back and take the long view. What's missing? What's not working? Revise and correct, and assign data updates with deadlines.

Define customer behavior keys. In this step you want to determine how you want to categorize behavior to enable statistical analysis. In other words, you may want to define customers by their buying track record, revenue level, profitability, etc. You'll need to identify both independent and dependent variables. This categorization will give you the tools to profile customers by certain "behavior" type, which is a critical component to customer modeling.

Create customer models. Based on the descriptive statistics developed for each customer, you're now ready to do some customer loyalty or revenue modeling, also called profiling. This effort enables you to begin to identify your "best" customers/prospects most likely to do business with you. Again, the more data you've collected on these folks, the tighter your message or offer will be.

Test and adjust. This portion of the process will be ongoing as your database matures. If you've built a solid, scalable marketing database, refinements and updates should be seamless and painless.

Overlaying and qualifying

Many people neglect what I consider a critical step in building a marketing database: prospect qualification. Purchasing a list from an outside source (see June 2000 "Lead Generation Then and Now" column or go to can quickly corrupt a good, clean house list. Therefore, it's important not to integrate these outside lists with your house list until it's been cleaned and qualified.

We work with a telemarketing firm in Southern California that's had lots of experience and success with graphic arts companies. They not only will call to confirm the accuracy of contact information, including finding the print buyer, but also will make appointments for sales. Call or email me, if you'd like their information.

Regardless of who conducts the qualifying calls, sales reps or telemarketers, it's a necessary step in lead generation and effective database marketing.

Charlotte Mills Seligman is president of Traversant Marketing Communications. The firm specializes in planning and executing integrated marketing programs for printing and allied graphic arts companies, with nearly two decades of expertise in the industry. Previous columns and issues of the company's Ti Monthly e-newsletter are posted on Inquiries should be directed to (415) 357-2929 or charlotte

© 2001 Charlotte Mills Seligman

May 20, 2001
Print & Graphics
Col#26, 5/01
Printing Journal
Col #26, 5/01
Prospect Databases
By Charlotte Mills Seligman

< Column Archives